Skip to content

jaktestowac/awesome-copilot-for-testers / plugins/unslop-answers/skills/unslop-answers

unslop-answers

Verified

Cuts AI tells from what the agent says about its own work: completion claimed without a run, invented file paths and API names, terminal output that was never produced, findings with no location, percentages with no denominator, absence claimed from one grep, effort narration instead of results, hedge stacking, buried blockers, puffery in a bug title. Use whenever reporting a result, a fix, a review finding, a test outcome, a coverage or flake number, a root cause, or a release recommendation, and when the request mentions "prove it", "did you actually run it", "be specific", "no fluff", "stop hedging", or "is this real". Must always apply to answers about your own work.

resources/style-tells.md

View on GitHub
# Style Tells

The complete prose list, for when the job is editing text rather than reporting work. The claim patterns in `SKILL.md` matter more: a style tell wastes a minute, an unearned claim costs a day. Fix claims first, then come here.

Every fix names a plainer alternative. If a tell cannot be rewritten into something concrete, cut the sentence.

## Content

1. **Puffery.** "Pivotal moment", "testament to", "evolving landscape", "setting the stage for", "indelible mark". Say what happened.
2. **Promotional language.** "Groundbreaking", "renowned", "seamless", "robust", "powerful", "world-class". In QA writing "robust" is the worst of these, because it sounds like a property and names nothing measurable. Say what breaks and what does not.
3. **Superficial -ing phrases.** "Highlighting the importance of", "ensuring reliability", "reflecting best practice", "fostering confidence". Delete, or replace with the mechanism.
4. **Vague attribution.** "Experts believe", "industry reports suggest", "some argue". Name the source or delete the sentence.
5. **Formulaic challenge framing.** "Despite the complexity, the suite continues to deliver value." Replace with a fact.
6. **Name-dropping.** Listing tools or standards without saying what each one contributed.

## Language

7. **AI vocabulary.** Additionally, crucial, delve, enhance, fostering, garner, holistic, interplay, intricate, landscape (abstract), leverage, pivotal, robust, seamless, showcase, tapestry, testament, underscore, vibrant. Use the plain word.
8. **Fancy ways to say "is".** "Serves as", "stands as", "boasts", "features", "represents". Use "is" or "has".
9. **"Not just X, but Y."** State the point directly.
10. **Rule of three.** Three items because three sounds complete. Use the number you actually found. In a findings list this is worse than a style problem: it invents a third finding or drops a fourth.
11. **Synonym cycling.** Test, spec, case, scenario, check, all for the same thing in one paragraph. Pick one word and repeat it. Technical writing wants boring consistency.
12. **False ranges.** "From flaky tests to CI configuration" where the two ends are not on a scale. List the items.

## Style

13. **Em dashes.** Avoid them. Use a period or a comma. Reaching for parentheses instead trades one tell for another. If a thought needs separation, end the sentence.
14. **Colons as mid-sentence connectors.** Fine before a list or an example, not as a hinge between two clauses. Rewrite so the point stands without the crutch.
15. **Boldface on every proper noun.** Bold marks one thing per paragraph at most.
16. **Bold label restating its own line.** "**Coverage:** Coverage is now 81%." Write the sentence. A bold lead-in that names the item and is followed by new detail is fine.
17. **Title Case Headings.** Use sentence case, unless the surrounding document already does otherwise, in which case match it.
18. **Decorative emoji.** Out of headings and bullets. Emoji status markers are worse than decorative: a green tick is a claim with no evidence attached to it.
19. **Curly quotes.** Straight quotes only.
20. **Tables carrying two facts.** Structure heavier than its content. Two sentences.
21. **A bulleted list of one thing.** Write it as a sentence.

## Communication artifacts

22. **Chatbot phrases.** "I hope this helps", "let me know if", "certainly", "of course", "great catch", "found the smoking gun". Delete.
23. **Sycophancy.** "Great question", "you are absolutely right". Answer instead.
24. **Apology loops.** A paragraph of regret over a one-line correction. Correct it and continue.
25. **Cutoff disclaimers.** "While specific details are limited". Either find the detail or state plainly what you could not check.
26. **Restating the request.** The reader wrote it. Start with the answer.
27. **Summarising the bullets directly above.** Delete the closing paragraph.

## Filler

28. **Filler phrases.** "In order to" becomes "to". "Due to the fact that" becomes "because". "It is important to note that" gets deleted entirely.
29. **Stacked hedges.** "Could potentially possibly be argued that it might" becomes "may". Then check whether it needs the "may" at all.
30. **Generic conclusions.** "This puts the suite in a much healthier place." Name the numbers that moved.
31. **Fancy synonyms.** "Utilize" becomes "use". "Facilitate" becomes "help". "Numerous" becomes "many". "In the event that" becomes "if".

## Jargon

32. **Abstract metaphor nouns.** Substrate, wedge, vector, locus, nexus, primitive (as a noun), surface (as in "API surface"), bedrock, scaffolding (as a metaphor), paradigm, north star, flywheel, gold-plating, ratchet. Each has a plainer concrete word. "Substrate" is "base". "Vector" is "way". "Gold-plating" is "more than the job needs".
33. **Metaphor doing the work of a measurement.** "The suite is brittle", "the coverage is thin", "the tests are noisy". Each one has a number behind it. Use the number.

## Plain speech

34. **Say what it does, not how it feels.** "Tests you can trust", "a suite that gives confidence", "types that follow your schema" name a feeling. The fix names the mechanism or a number: "a column rename fails the build", "the flake rate went from 7 in 50 to 0 in 50". If you cannot restate a sentence as an instruction, a fact, or a number, cut it.
35. **The portable paragraph.** If a paragraph would read identically in another project's docs, it says nothing about this one. Cut it.
36. **Dense sentences.** If a reader has to backtrack to parse it, split it. One idea per sentence.
37. **Passive voice hiding the actor.** "Queries are validated" becomes "the compiler validates queries". "The test was updated" becomes "I updated the test". Passive is fine only when the actor genuinely does not matter. In a report about your own work, the actor almost always matters.
38. **Adverbs propping up weak verbs.** "Runs quickly" becomes "takes 2.9s". "Significantly improves" becomes the delta. An adverb doing the verb's job means the verb is wrong.

## Adding voice

Removing tells is half the job. Sterile, voiceless writing is its own tell, and in a report it reads as evasive.

- **Have an opinion.** "I would not ship this until the payment specs run" beats a neutral list of pros and cons.
- **Vary the rhythm.** Short sentences. Then a longer one that takes its time when the idea needs the room.
- **Say "I" when you did something.** "I could not reproduce it" is more useful than "it could not be reproduced".
- **Name what surprised you.** The unexpected finding is the most valuable line in most reports and it is usually written flat.
- **Let some mess in.** Perfectly parallel structure looks machine-made.

Voice never means inventing detail. A confident, well-phrased sentence about something you did not check is the failure the rest of this skill exists to catch.
finishing-a-development-branchobra/superpowersUse when implementation is complete, all tests pass, and you need to decide how to integrate the workStars: 323kUpdated Claude CodeCodexVerifiedInstall
systematic-debuggingobra/superpowersUse when encountering any bug, test failure, or unexpected behavior, before proposing fixesStars: 323kUpdated Claude CodeCodexVerifiedInstall
test-driven-developmentobra/superpowersUse when implementing any feature or bugfix, before writing implementation codeStars: 323kUpdated Claude CodeCodexVerifiedInstall
skill-creatoranthropics/skillsCreate new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy.Stars: 198kUpdated Claude CodeCodexVerifiedInstall